Independent Assessment of Housing Needs Research

The full document is published on the JPU website and can be download by clicking here (this will open in a new window).

The text of the document can also be found in the sections below

Determining Objectively Assessed Housing Need (OAHN) totals for West Northamptonshire and the Constituent Local Authorities – a Commentary






Report for West Northamptonshire Joint Planning Unit

December 2013

Richard Cooper (BSC MRTPI)

Determining Objectively Assessed Housing Need (OAHN) totals for West Northamptonshire and the Constituent Local Authorities – a Commentary




Commissioned work


          A. The 2011-based household projection as it stands
          B. Reworking of population element(s) of the projection
          C. Reworking of household elements of the projection



Appendix 1: Extract from ONS Local Insight Reference Panel, Manchester, 17 July 2013

Appendix 2: Migration Age profile for 2001-11 PBA projection supplied to WNJPU by ONS

Appendix 3 – Consideration of the conclusions of JPU report

Appendix 4 – Additional note on unattributable change in ONS estimates






Report for West Northamptonshire Joint Planning Unit

December 2013

Richard Cooper (BSC MRTPI)


1.   Two pieces of work were commissioned by the JPU to assist with the assessment of local generated housing need for West Northamptonshire and its constituent local authorities.

2.   The results of these pieces of work suggested a number of options with a range of assumptions, giving various levels of projected housing. The JPU subsequently has produced a report drawing conclusions from those reports[1] . These three reports are already available and will be referenced in this report. It is not the purpose of this piece of work to repeat that work or provide any further detail or analysis[2] . Anyone wishing to study in detail may return to those reports.

3.   The purpose of this report is to provide an independent overview of those three reports and the conclusions drawn in the JPU report, in the context of current advice from Government, plus the Local Government Association, Planning Advisory Service and ‘

4.   This report recognises that the JPU has undertaken to meet the stated aims of the National Planning Policy Framework that Local Plans should meet objectively assessed housing needs, that local planning authorities should have a clear understanding of housing needs in their area and should meet household and population projections, taking account of migration and demographic change.


5.   The 2008-based ONS & CLG projections, which until a few months ago would have been the latest projections used as a starting point for establishing an OAHN, have been superseded in two principal respects, although other concerns are expressed in Para 3.2 of the JPU report.

  • The Census has provided a new (and for WN, lower) population base on which to establish a projection. As a result it has or will also provide data for:
    • a) new fertility and mortality rates (not yet available)
    • b) Migration rates (not yet available)
    • c) Household representative rates (partially available)
  • The Government has also prompted the production, in April 2013 of interim part-census-based projections of population, and consequently households. However, these only run to 2021, requiring further work to extend them to 2031.

6.   In his consideration of the Joint Core Strategy the JPU Examination Inspector indicated the need for further work to take into account both the latest Office for National Statistics (ONS) housing projections and the early 2011 Census results.

7.   As stated in Para 3.2 of the JPU report:

“The JPU considers that there is a high degree of uncertainty with the preparation of population and household projections at the current time, and all the indications are that this will not change until the 2012 based Sub National Population Projections and associated Household Projections are published in Spring and Summer 2014.”

Commissioned work

8.   Two consultancies were asked to advise the JPU on the objectively assessed housing need for the area. They each approached the work differently with proposals as described in their reports. It should be stressed here that this is not a critique of that work; rather it is a commentary on their application to a plan that needs to meet the requirements of the National Planning Policy Framework, including establishing an OAHN and suitable housing requirement in this context. Both pieces of work undertook valid and exemplary approaches to what they were asked to do. I have restricted my commentary in the main to the JPU area as a whole.

9.   In brief the consultants reported as follows:

  • The Peter Brett Associates report highlights population change 2001-11 as the major factor affecting the veracity of both the 2008-based and the 2011-based household projections. The report describes two of their own trend based population projections, based on the 2011 official projections and 2001-11 estimate components of change, to provide a projection for 2011-2031. From these it suggests a minimum housing requirement. No change is made to the CLG’s household formation rate trends.
  • The Cambridge Centre for Housing and Planning Research paper identifies the difference in household change to be significant, and has used the 2011 Census based population projections unchanged, but taken a different approach to projecting the official household projections forward to 2031.
  • The paper suggests two alternative modifications to household representation rates for 2015 – 2031 to increase the 2011 based household figures to what it considered to be more appropriate level (part return to trend and 2008 tracking). These two numbers have very similar results, and are also similar to the JCS figure.
1. [back]
2. With one exception, since much of the work described here has been done ONS has released rolled forward mid-2011 population estimates which give an indication of the differences between the Census-based estimate and those based upon the 2001-10 methodologies. This supersedes the use in the studies of surrogate 2008-based projection totals for 2011. [back]


10.   This report goes on to consider the suitability of:

A. The 2011-based household projection as it stands;
B. Reworking of population element(s) of the projection;
C. Reworking of household elements of the projection.

This in turn relates to the methodologies and robustness of the data elements used in updating the projections, and how suitable any updates are when applied in the JPU Plan Area, which due to recent planning history has some particular characteristics.

A. The 2011-based household projection as it stands.

11.   The PBA report suggests that the 2011-based Population projections inadequately address the matter of updating the projections in methodological terms[3] .

12.   While being based upon a 2011 Census-based population estimate, the interim population projections have utilised fertility, mortality and migration data obtained from population estimates & projections that pre-date the Census. Consequently there are concerns that the application of such rates to a Census-based population which may differ in age-sex profile has produced results that will vary from those which, in due course, will be produced using compatible data.

13.   With regard to migration it should be pointed out that while the various ONS population projections referenced in the reports are referred to as ‘trend-based’ they do not project forward on the basis of an average past net migration[4] . Thus, while on the face of it a lower updated population change would imply a lower migration level and projected population the relationship is not straightforward. This is especially true when derived fertility, mortality and migration rates are being applied to a different age structure to that which generated them.

14.   The population age structure of the 2011 population estimate could have an impact upon the projections. This would depend upon the difference between the 2010 estimates and the 2011 Census-based ones. If there were little difference then there would be fewer concerns about projections being less reliable. One example identified in some locations is the different age structure for young women upon which an inappropriate fertility rate has been applied. This is mentioned in the reports [PBA 3.15, CCHPR 3.8], however a full analysis has not been done to my knowledge.

15.   Since much of the work described here has been done ONS has run a series of workshops to introduce the comparison of mid-year estimates based upon the 2001-10 methodologies. The data produced indicates that the rolled-forward (based on 2001-10 methodologies) estimate for 2011 would be between 5,000 & 8,000 higher than the Census-based estimate. This is 1.7 - 2.1% different.

16.   The published comparison of estimates from ONS shows that for West Northamptonshire there is a significantly lower 18-30 male population in the 2011 Census-based estimate compared to the 2001-2010 methodologies. (Males & females 39-45 also show a slightly lower figure). The differences are concentrated in Northampton. This suggests that the 2008-based projection inflated migration owing to the age structure of the (2008) estimate having more total population, a higher propensity to migrate through past change and a greater population in that mobile age group.

17.   The West Northants age structures are replicated in Appendix 1.

18.   With the Household projections matters are somewhat more straightforward, although the differences seen in the 2011-based household projection compared to the 2008-based one, as a result of the Census, show significant reduction in household generation. The methodology used by CLG is little changed from the earlier projections, the differences and trends are clear, and the data has not been applied to an inconsistent base. The work appears to be robust and well-received.

19.   Conclusion: I would agree with the reports that for West Northamptonshire[5] the 2008-based projection is not a valid basis for setting a OAHN, and is too high to be used as it stands. The 2011-based interim projection has doubts in terms of its methodology but is more realistic. It does display an uplift in population growth from 2001-11 that appears to track, from a lower base, the 2008-based growth [PBA fig. 3.3]. However, because complete 2001-11 change data has not been used in its production, doubts remain as to whether that is an effect that will continue in future estimates or projections.

20.   Concerns over the household projections revolve around whether the trends seen in household representation (reflected in a slow-down or turn-round in the reduction in household size) will continue. These relate less to methodology, rather how past trends are applied to future rates to give long-term trends, and are discussed below.

B. Reworking of population element(s) of the projection;

Population – PBA report

21.   The report further proposes updated projections, with births, deaths and household formation using the same assumptions as the official 2011-based projections, but migration rates being adjusted to reflect post-Census (total) population change. These projections take two periods, 2001-11 and 2006-11. I understand that they are not controlled to meet an average net total migration level; rather that they operate in a similar way to the ONS projections. Hence the comments in the appendix that average net migration 2011-31 differs from 2001/6-11 [PBA app. p22].

22.   The PBA report projections were based upon migration data for two periods: 2001-11 & 2006-11, with the later period showing a lower net migration and producing a lower projected population. The impact of the Grange Park development on growth in South Northamptonshire before 2006 is remarked upon [PBA 4.18]. The charts demonstrate that much more significant for the JPU Area is the increase in Northampton’s population and migration 2005-7. This overlapped with South Northamptonshire’s growth, but for the Plan Area as a whole growth peaked in 2005/6, after South Northamptonshire’s growth had peaked [PBA figs 4.4 – 4.6].

23.   What was driving this growth appears to have been inter-regional and international migration, and housing completions followed by population change [fig 4.6 & PBA Appendix]. Consequently notwithstanding the effect of Grange Park in the early part of the 2001-11 period, the post-2007 decrease in net migration is likely to be attributable as much to inter-regional and international economic factors (e.g. labour and housing market). In any event this emphasises the conclusion in the PBA report that its lower 2011-31 provision figure should be a minimum [4.11 – 4.13].

24.   Turning to the PBA 2001-11-based figure this is based upon a longer trend period and as the report indicates [3.22] this would usually be considered a better basis for long-term projections. I would concur with PBA that this would usually be considered a better basis for long-term projections. The resulting projection gives a higher population. Also, the projections are based upon migration age profiles for each LA (Appendix 2), which are very different, for example, Northampton attracts more younger people with the other two LAs losing them (net). Bearing this in mind the effect of a swing from South Northamptonshire to Northampton, and an increasing impact of international migration in the late 2000’s would affect the way in which any projection would develop. This also supports a longer period to establish the migration profile in that it incorporates a broader set of circumstances.

25.   There is one caveat to this view; the PBA Appendix indicates that annual average net migration from the projections 2011-31 is below that of the 2001-11 period (after Census adjustment)[6] . A projection set to maintain net migration at 2001-11 levels would, therefore, produce a higher population (assuming natural change effects to be the same).

26.   Conclusion: PBA have adopted an approach to population which is technically and methodologically soundly based, i.e. revising the population projections behind the CLG household projections on the basis of more consistent data (but incidentally not altering the household representative rates). However, it is more subject to uncertainty and the application of trends, particularly in migration, that it may not be advisable to apply. PBA recognise that the approach indicates a minimum figure for an OAHN.

27. Also borne in mind is the context in which the work is to be used. The factors set out in this section would indicate that the migration informing the PBA projection produces an outcome that may under-estimate future population growth. These derive from the nature of the projection in not being controlled (par. 25) and the trends that inform it.

Population- CCHPR Report

28.   In contrast to the PBA report the CCHPR report states:

“The 2011 census (on which the 2011-based interim population projections are based) found 3.3% fewer people in the West Northamptonshire JCS area than the 2008-based projections had projected. However, both projections envisage similar population growth – an increase of 45-46,000 or 12% over the period 2011-21. The difference between the two projections for the increase in the number of people in the area is only 550 people or 1.2% of the projected growth. It should, however, be noted that there are changes to the projected births, deaths and internal and international flows into and out of the three authorities – and that these changes broadly cancel each of other out with the result that the overall projected population growth for the JCS area is little changed.” [CCHPR Executive Summary]

29.   The CCHPR report also notes that in comparison the difference in the household projections is much greater (almost 20%). CCHPR thus conclude that the 2011-based interim population projections are a satisfactory basis for their work and do not undertake a separate projection. The 2011-based interim projections are extrapolated from 2021 in 10-year age groups on a simple basis [4.4-on]. CCHPR considers that this more straightforward method is satisfactory in establishing a 2011-31 population change that serves the purpose of its work. Consequently the population basis of the work is effectively that of retaining the 2011-based projections with their associated doubts (see above ).

30.   Conclusion: The CCHPR work appears to take a proportionate approach to the need to review population growth, recognising the scale of the differences in projection. It also has the advantage (in terms of meeting CLG guidance) of using the Official (2011-based) projections up to 2021.

31.   Conclusions on Population: The different conclusions drawn in each of the reports derive from, for PBA, a consideration of the technical background to the 2011-based projections, and for CCHPR, how much impact did the updated projections have in this case? While for other areas the latter approach may not necessarily be satisfactory, it appears to be so for the JPU area, because of the limited impact of the Census on population figures.

C. Reworking of household elements of the projection;

32.   The Household elements of the projections are more straightforward, and less subject to uncertain local circumstances and changes in trends which are then applied to the future. The CCHPR and WNJPU reports describe the issues and process very clearly. The decisions taken are more judgment-based than technical, and related to whether the future trends (in household composition/ representation) will be affected by economic & social factors, rather than whether the identified trends are technically accurate.

33.   The PBA approach is described in their appendix:

“The conversion to households uses the household representative rates and other assumptions of the CLG interim 2011 projections ... to 2021. After 2021 the household representative rates from the CLG 2008 projection are used with gender/age/relationship adjustments based on the comparison of rates with the CLG 2011 projection at 2021.”

34.   In other words they continue the trends established in the official projections.

35.   In contrast the CCHPR report considers that because the difference in growth between the 2011-based and the 2008-based trend in households is so great (20%), then it is “the household formation rates that deserve close attention”. Consequently the report summarises the issue as:

“... whether those lower rates are indicative of a new longer term trend that will be maintained into the future or merely a short-term departure from the underlying trend caused by the economic downturn and a sustained period of poor house price affordability.”

36.   The principle adopted by CCHPR is that the trend of reducing household size seen over a long period (the CLG projection uses 5 decades of data) will be returned to in some way in due course, with the effect of increasing the 2011-based household projection total at 2031 by 4-5,000 (appr. 2.5%).

37.   In technical terms the work is quite straightforward and the uncertainty associated with future scenarios by no means as complex as with the population projections. The main doubt relates to further Census data, yet to be published, that will refine, rather than re-base, the household representation rates.

38.   The report’s discussion about the aspects of future change in two age groups make assumptions about possible futures.

  • That 25-34 year-olds will return to the long-term trend of increasingly heading households (partly due to living less with parents).
  • That for 60-74 year-olds the more recent trend is somewhat more likely to continue, but in part.

39.   Adopting an altered household representation rate appears to be justified. It is a pragmatic approach to a situation that CCHPR describe thus:

“By following one or other of the (virtually indistinguishable) partial return scenarios the risk of needing to make a significant adjustment would be reduced, although regular monitoring of what actually happens would still be important”

Further and more detailed study of projected household change is included in CCHPR’s report for the Town & Country Planning Association which comprises the T&CP Tomorrow Series Paper 16: New Estimates of Housing Demand and Need in England, 2011 to 2031. That report proposes a ‘modified trend projection’ which corresponds largely to the ‘partial return to trend’ used in their (earlier) study for WNJPU.

40.   Conclusion on households: The lack of any alteration to household representation trends is an aspect of the PBA work that I would not support; the application of the uplift in household change indicated in the CCHPR work (2.5%) could be applied to the PBA outcomes to indicate an approximate minimum figure that took account of changes in household formation alone. As CCHPR indicate this is a situation that should be monitored, particularly with Census results in 2014 providing more detail. I would suggest that the approach taken by CCHPR with the ‘partial return to trend’ is more sensible. While it is pragmatic rather than technically correct it represents a circumspect approach which would take account of a potentially faster recovery in household formation and greater housing need. Having said that, I acknowledge that the difference between the two CCHPR alternatives is small, and there are much greater potential impacts on future household change arising from other sources.


3. The full comment is worth repeating: “The 2011-based official projections take account the Census results but only partially. They start from the correct population at 2011... But the migration rates that the new projections roll forward are taken from 2010 estimates, which predated the Census. This is a serious flaw, because the Census shows that these estimates were wrong – though in the case of West Northamptonshire they were nearer the truth than the 2008 ones.” [back]
4. The projections take a representation of a migration age profile (or propensity to migrate, both in- and out) typical of the previous five years, and apply that to the first year of the projection. Subsequent years continue to apply that profile & consequently any net migration result will vary over time. [back]
5. NOTE: this conclusion could not necessarily be drawn for other areas where differences may be greater. [back]
6. “The 2001-11 Trends projection results in natural change averaging 2,000 and a net migration gain averaging 1,000 per year between 2011 and 2031, compared to the averages of 1,700 and 1,300 in 2001-11.” PBA Appendix p22) [back]


41.   The question still remains as to whether trend-based population projections as produced by ONS and PBA represent a likely future bearing in mind the recent past and impact of planning strategies. This question is addressed in the guidance produced for by CCHPR which cites the impact of atypical past events as one reason not to adhere to the ONS trend-based projections[7] . I would suggest for the reasons set out in par. 22-28 above that the same could apply to the PBA projections. The PBA projections should therefore be utilised with recognition of their technical basis, and this supports PBAs comment that they should be treated as a minimum.

42.   CCHPR, in accepting the 2011-based population element, has, for the West Northamptonshire context at least, adopted a ‘low-risk’ approach less sensitive to anomalies within this Plan Area.

a) The differences in growth rates between the 2008 & 11-based population projections appear, on the face of it to be small. This implies that, while applying pre-census fertility, mortality and migration data to a Census base is methodologically unsound, the impact for WN appears to have been minimal. However – this applies only to total population & any anomalies relating to age structure in projecting into the future may be hidden on this analysis.

b) The use of a post-census migration age structure appears unlikely at this stage to have a great impact upon the projection.

43.   The scale of net migration implied in the PBA work utilising the post-Census migration age structure, however, has a significant impact, reducing growth to match 2001-11 rates as altered by the Census results. The nature of the projection means that this outcome is not controlled to 2001-11 net migration, and is lower than those levels.

44.   With regard to household formation, some adjustment is warranted, although the potential for higher, rather than lower household generation is more likely, a situation that should be monitored.

45.   When set in the context of the HPPF and an OAHN the PBA work, while valid in its own right, and to some extent replicating the ONS projection, may not be appropriate to the task of identifying the OAHN, as acknowledged by PBA[8] . Adopting a higher level of housing (than the PBA figures) is precautionary (for housing delivery) bearing in mind doubts over future migration expressed above. It also incidentally reflects the particular concern over low South Northamptonshire figures raised in the WNJPU report, and gives scope to resolve that outside the technical consideration, i.e. as a whole HMA/3 authority planned approach.

46.   Whichever approach is taken the results only represent the starting point for establishing an OAHN. As quoted by PBA, ‘ comments, in relation to the CLG household projection:

“The temptation to jump to the conclusion that this is “the answer” for the number of homes that need to be built each year should be resisted. It is the figure that would be needed to house new households if recent trends were to continue. This is a starting point for further analysis and discussion, not an answer.”

47.   The term ‘recent trends’ also is not straightforward, as these require understanding and context, which the two commissioned reports have provided, and this report attempts to supplement. Each report makes this clear [PBA3.4,3.5, CCHPR 3.9].

Consideration of the conclusions of JPU report

48.   A direct response based on the JPU report conclusions is contained in Appendix 3.

7. Choice of Assumptions in Forecasting Housing Requirements: CCHPR (March 2013) – Note 3, Para 6, [back]
8. In correspondence PBA have stated “an NPPF-compliant target cannot be lower than this [suggested] number. But it may be higher, even much higher.” [back]

APPENDIX 1: Extract from ONS Local Insight Reference Panel, Manchester, 17 July 2013

Statistics for West Northamptonshire

(comprising the districts of Daventry, Northampton and South Northamptonshire)

3. Age-sex comparisons

The following graphs show, for each sex, the age distribution of the population in:

i) The official (Census-based mid-2011) estimates.
ii) The old method rolled-forward mid-2011 estimates.
iii) The rolled-forward mid-2011 estimates plus the attributable components of difference.

This shows how much of the difference between the Census and rolled-forward estimates was explained by the unattributable factors, and how much was allocated to Other.

RC App1_F1
Figure 1. The official (Census-based) and old method rolled-forward mid-2011 population estimates for males in West Northamptonshire, plus the old method rolled-forward mid-2011 estimates including the attributable components of difference.
RC App1_F2
Figure 2. The official (Census-based) and old method rolled-forward mid-2011 population estimates for females in West Northamptonshire, plus the old method rolled-forward mid-2011 estimates including the attributable components of difference.
APPENDIX 2: Migration Age Profile for 2001-11 PBA projection supplied to WNJPU by ONS

RC App1_F3
Persons by age

APPENDIX 3 – Consideration of the conclusions of JPU report

The JPU’s conclusions are set out in the paragraphs below.

My response to those conclusions are set out in italics.

Extract from West Northamptonshire Joint Core Strategy - Objectively Assessed Housing Needs - December 2013 (Text only, tables removed)

8.5   The Peter Brett Associates projections should be discounted, as they are considered to be too low compared to the latest official projections (2011 based Interim Projections) and do not therefore constitute the basis for a positively prepared plan in accordance with the requirements of the NPPF.

The PBA projections may be discounted for the second reason given but not, per se, because they are ‘too low’ However, I consider them to represent an unreasonably low scenario to encompass what would be an OAHN, in both population and household aspects.

8.6   The 2008 DCLG Projections should be discounted as being too high. This data set has been largely discredited by the results of the 2011 Census, especially in the West Northamptonshire area where the 2011 Census results were demonstrably below the 2008 based projections; a position reinforced by the Cambridge Centre for Housing and Planning Research advocating use of the 2011-based information for West Northamptonshire over the 2008-based data behind the “What Homes Where” Toolkit.

There is general consensus that this is the case.

8.7   The remaining projections are all very similar in the scale of development they suggest. However, the JCS projections should be discounted as they were produced on a dwelling constrained basis, and therefore do not comply fully with the currently understood definition of Objectively Assessed Housing Need.

It would not be appropriate to maintain a reliance upon the JCS projections as they were based on superseded data.

8.8   Of the two CCHPR projections, the one which it is recommended by the Joint Planning Unit to form the basis of the Objectively Assessed Housing Need for West Northamptonshire is the Tracking projection. This has been selected on the basis of the description of the methodology in Paragraph 6.1 of the CCHPR report.

8.9   The CCHPR report admits that there is no particular science behind the alternative “Return to Trend” option as the “half-way” assumption has no real justification other than it is unlikely that there will be no move back towards trend and improbable that there will be a full return in the foreseeable future. On the other hand, the Tracking methodology which is being recommended, is based on the difference between the 2008 based projections for 2011 and the actual result as expressed in the 2011 Census, and this difference being maintained throughout the period of the CCHPR “Tracking” projection.

The approach taken by CCHPR with the ‘partial return to trend’ is in my view most sensible. It is pragmatic rather than technically correct and it represents a circumspect approach which would take account of a potentially faster recovery in household formation and greater housing need. Nevertheless the difference between the two CCHPR alternatives is small, and there are much greater potential impacts on future household change arising from other sources. Consequently I would accept the conclusion of the JPU in these circumstances.

8.10   The Joint Planning Unit considers that the table below sets out the Objectively Assessed Housing and Jobs Needs for West Northamptonshire: (Table omitted)

8.11   The only reasonable alternative is the extended 2011 Census-based projections shown below, and described more fully in paragraphs 4.9 to 4.11 of the CCHPR paper. This would give totals lower than the CCHPR totals set out in Tables 7 and 8 above, but do represent a projection of the latest official Government projections. These figures also represent a significantly higher figure than the PBA recommended totals, and PBA do indicate their belief that the 2012 based SNPP projections, due to be published in 2014, will be lower than the 2011 Census-based projections for West Northamptonshire. The 2011 Census-based projections could therefore be seen to represent an over-provision against what is expected to be in the next set of official projections, and could therefore be argued would represent positive planning for the area. The relevant figures are set out in Tables 9 and 10 below: (Table omitted)

The extended 2011 Census-based projections only represent the latest Official projections to 2021, and they were undertaken for other than long-term planning purposes. However, in that context the conclusions and comments of the NWJPU are valid. It should be noted that the next set of projections will be, in the same way, trend-based.

8.12   The JPU does not consider there are any other reasonable alternatives. For the reasons given in the report above, the PBA recommendation is considered to be too low and not represent positive planning for the area, and the 2008-based Sub National Population Projections are considered to be too high, as evidenced by the results of the 2011 Census. The alternative CCHPR totals and the extended JCS totals are too close to the recommended Objectively Assessed Housing Needs totals to be reasonable alternatives.

Re-iteration; No comment

APPENDIX 4 – Additional note on unattributable change in ONS estimates

I have been asked to comment on an unattributable change component in the 2002-11 revised Population Estimate data (of about minus 800 p.p.a) for all three authorities. ONS describe this thus:-

“Taking all attributable components of difference into account, some LAs are left with a small Other (unattributable) component whereas others have a large Other component, sometimes larger than the original difference. This may be because of inaccuracy in the attributable components, most of which involve estimation and assumptions. However, it is also likely that other, unattributed factors have made a substantial contribution to the difference. “[9]

This ‘Other’ component has to be almost wholly attributed to migration.

It is noticeable that the rates of change shown in Section 4 of the PBA report (fig 4.1 on) show that past population change in the area, over the period sourced for the official projections and those in the PBA, has seen significant variation, especially before & after 2007.

Therefore apportioning this, as ONS has done, equally across the period may ignore the fact that it relates to a certain part of the period (for example the early period when Grange Park was being developed). I recognise the concerns expressed by the JPU; for PBA to change how this component was accounted for would make a difference to the 06-11 projection, but is not likely to affect significantly any projection based on the 2001-11 period.

It should be noted that this unattributable change, by its nature, is unlikely to be included in migration components in the future, either for estimates or for projections. Its effect may be to increase the level of migration in the 2012-based projections when they are produced. This is speculative and does not affect the conclusions of this work, however.

While I have not considered LA-level projections as such, it appears clear that using the 2006-11 migration trend over-emphasises a growth in Northampton against a very small growth in South Northamptonshire. This may arise from the impact of the ‘Other’ change mentioned above, which if it arose from inaccuracies prior to 2006 due to the Grange Park development, would address the low growth in South Northamptonshire. However, any future provision of housing within the JPU Area is more a matter of suitability and planned provision than historical demographic change.

9. ONS Report - See for more information [back]